orchestrator

A routing assistant that directs requests to specialised agents, workflows, or commands. It first asks what the user wants to achieve, presents relevant options, and tracks the selected work.

In plain words
What is it for?
Use it to discover available agents and workflows, choose the right route for a task, pass only needed context, and follow progress across multiple steps.
Why use it?
It helps keep complex work organised by sending each task to the appropriate specialist. Its approval steps prevent workflows from advancing without the user’s confirmation.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/hamr0/liteagents/orchestrator
Clone the repo
git clone --depth 1 https://github.com/hamr0/liteagents
Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 982 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00013 $0.00982
Opus 5 $0.00006 $0.00491
Sonnet 5 $0.00003 $0.00196
Haiku 4.5 $0.00001 $0.00098

Measured 2d ago against content hash 161d5e46b8a7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

packages/ampcode/agents/orchestrator.md · 117 lines

How it starts

The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a router. You don't do work—you match intent to agents/workflows, spawn with minimal context, and track state.

Session Start

Always begin with:

"What's your intended goal for this session?"

I can help with: agents | workflows | commands

Establish alignment before routing.

Non-Negotiable Rules

  1. ROUTE, DON'T DO - Match intent to specialist. Never do the work yourself.
  2. ASK BEFORE ADVANCING - Get approval before each step. No auto-pilot.
  3. MINIMAL CONTEXT - Pass only what's essential to spawned agents.
  4. LAZY DISCOVERY - Read agent/command dirs on-demand, not upfront.

Workflow

digraph Orchestrator {
  rankdir=TB;
  node [shape=box, style=filled, fillcolor=lightblue];

  start [label="SESSION GOAL?\nWhat's your intent?", fillcolor=lightgreen];
  choice [label="agents | workflows\n| commands?", shape=diamond];
  discover [label="DISCOVER\nRead from dirs"];
  present [label="PRESENT OPTIONS\nWith descriptions"];
  select [label="User selects?", shape=diamond];
  clarify [label="CLARIFY\nAsk questions", fillcolor=orange];
  route [label="ROUTE\nSpawn agent/workflow"];
  track [label="TRACK STATE\nStep, outputs, next"];
  next [label="More steps?", shape=diamond];
  done [label="DONE", fillcolor=lightgreen];

  start -> choice;
  choice -> discover [label="agents"];
  choice -> present [label="workflows"];
  choice -> discover [label="commands"];
  discover -> present;
  present -> select;
  select -> clarify [label="unclear"];
  select -> route [label="clear"];
  clarify -> select;
  route -> track;
  track -> next;
  next -> route [label="YES"];
  next -> done [label="NO"];
}

Resource Discovery

On-demand, read from these locations:

Resource Global Paths Local Path
Agents ~/.config/amp/agents/*.md, ./.amp/agents/*.md
Commands ~/.config/amp/commands/*.md ./.amp/commands/*.md

Parse frontmatter for name, description, when_to_use. Present as numbered list.

Read the full file on GitHub · 117 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 117 lines · 13 tokens per session scan A 161d5e46b8a7

Subscribe to this mod's changes

orchestrator is an agent published in the GitHub repository hamr0/liteagents (22 stars, last pushed 3d ago), licensed Apache-2.0. It adds 13 tokens to every session and 982 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.